한국어 전산통사분석의 여러 양상 -한영, 영한 기계번역 시스템을 위한 경우-

Abstract

This is a brief survey of some considerations necessary in parsing Korean in terms of computation. The parser-generator used was part of ECS Machine Translation Engine which is still being upgraded. First, the general scheme of UKOG (my Korean grammar named Unification-based Korean Grammar) is presented and then discussed are five areas of consideration which are deemed to be important for processing Korean; namely (1) the level of ambition of your goal, (2) written vs. spoken data, (3) linguistic aspects, (4) computational aspects, and (5) exchange of information among related disciplines. Especially the third section (3) linguistic aspects (see 5.3) focuses on 12 topics which I consider are the major important considerations that are more or less specific to Korean.